Preconditioning constrained eigenvalue problems (Q1025859): Difference between revisions

The purpose of the paper is to introduce a robust preconditioning scheme for the numerical solution of the leftmost eigenvalues and corresponding eigenvectors of the constrained eigenvalue problem \[ \left[ \begin{matrix} A & B^T\\ B & 0\\ \end{matrix} \right] \left[ \begin{matrix} u\\ v\\ \end{matrix} \right]= \left[ \begin{matrix} M & 0\\ 0 & 0\\ \end{matrix} \right] \left[ \begin{matrix} u\\ v\\ \end{matrix} \right]\lambda \tag{1} \] \(\;A,\,M=M^T\in\mathbb R^{n\times n},\,B\in\mathbb R^{m\times n},\,m<n\) The first section represents an introduction in nature. The second section focuses on the structure of the eigenvalue problem. Thus, the authors demonstrate that even if \(A\) and \(M\) are symmetric matrices, and in addition that \(M\) is positive definite, (1) is not a generalized positive definite eigenvalue problem. This constrained eigenvalue problem is congruent to a nonsymmetric eigenvalue problem with nontrivial Jordan blocks associated with infinite eigenvalues. The third section presents the preconditioning of the constrained eigenvalue problem. The proposed preconditioning scheme is relevant to the application of Krylov subspace methods and preconditioned eigensolvers. The two key results are a semi-orthogonal decomposition and a transformation process that implicitly combines a preconditioning step followed by abstract projection onto the subspace associated with the finite eigenvalues. Two numerical experiments which demonstrate the effectiveness of the preconditioning scheme are presented within the fourth section. The last section is devoted to the main conclusions and the future work.